With　the　rapid　development　of　biology,　its　relevant　literature　will　be　more　and　more　important　for　researchers　to　dig　out　more　valuable　knowledge.　Named　Entity　Recognition　(NER)　for　biology　literature　is　a　very　common　and　important　task　in　these　works.　With　the　increase　of　the　literature　amount,　the　recognition　speed　becomes　slower　and　less　accurate.　In　order　to　tackle　this　problem,　a　MapReduce　based　NER　is　used　to　reduce　the　running　time　by　leveraging　parallel　processing.　Experimental　results　show　that　the　application　of　entity　naming　recognition　to　the　MapReduce　framework　has　greatly　improved　in　the　CRF　model　training　process　and　entity　recognition　process.　Name　entity　recognition　is　accommodated　to　parallel　framework　to　solve　recognition　problem　rapidly　and　accurately.　©　2017　IEEE.